Mutations in any one of three distinct genes can cause early-onset Alzheimer disease (AD) that is similar to the much more sporadic form that occurs most frequently in older individuals. An invariant feature of AD pathology is the accumulation of amyloid plaques in the brain, composed primarily of the highly amyloidogenic pathology is the accumulation of amyloid precursor protein (APP). Elevation of long-form Abeta peptide is observed in presymptomatic carriers of disease-linked mutation in APP, presenilin 1 (PS1) or presenilin 2 (PS2) suggesting that APP processing is central to the disease process. The identification of genes and mutations causing AD have made possible the development of transgenic mouse models that recapitulate aspects of AD, including development of amyloid plaques and behavioral abnormalities. More importantly, the biochemical pathways involved in D seem to be similar in mice and humans; for example, mutant but not wild-type,. PS1 transgenes cause elevation of Abeta1-42 derived from a human APP transgene or from the endogenous mouse APP gene. Genetically defined stocks and strains of mice offer a new avenue for exploration of AD. Age of onset has been shown to vary considerably in large families segregating a single FAD mutation suggesting the existence of genetic modifiers, and sporadic AD almost certainly has a genetic component. By exploiting natural variation among different inbred mouse strains, genetic control of susceptibility to effects of APP over-expression on survival, pathology, behavior and electrophysiology. The ultimate goal is to identify the genes involved. By transferring the Tg2576 transgene-array from the outbred stock that develops amyloid plaques onto different inbred backgrounds, it will be possible to determine the relationship of plaque load to disease phenotypes. Preliminary genetic analysis indicates the presence or absence of behavioral abnormalities in mice with amyloid plaques depends on genetic background rather than plaque prevalence. Testing the Tg2576 transgene array in hybrids with recombinant inbred strain panels will allow comparison of the genetics of complex behavioral and electrophysiological traits that require replicates measurements on genetically identical animals. These and similar experiments will help determine whether extracellular amyloid is essential for the cognitive decline and neuronal loss seen in AD. Finally, testing the effects of candidate genes, including apolipoprotein E and superoxide dismutase 1, on the phenotypes produced by APP over-expression provides a means to dissect disease pathways and refine mouse models for AD.